AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

Bridging the Skeleton-Text Modality Gap: Diffusion-Powered Modality Alignment for Zero-shot Skeleton-based Action Recognition

Jeonghyeok Do, Munchurl Kim

ICCV 2025 · Vision-language and multimodal › CLIP and zero-shot · approach: Attribute-based zero-shot learning · 10 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add triplet diffusion loss to the diffusion-loss-only model (ablation)TDSM · NTU-60 55/5 splittop-1 accuracy (%): 79.87 -> 86.49Improved
add triplet diffusion loss to the diffusion-loss-only model (ablation)TDSM · NTU-120 96/24 splittop-1 accuracy (%): 57.65 -> 65.06Improved
add diffusion loss to the TD-loss-only model (ablation)TDSM · NTU-60 55/5 splittop-1 accuracy (%): 80.90 -> 86.49Improved
use both global and local text features instead of global only (ablation)TDSM · NTU-60 55/5 splittop-1 accuracy (%): 83.41 -> 86.49Improved
use both global and local text features instead of local only (ablation)TDSM · NTU-120 110/10 splittop-1 accuracy (%): 69.95 -> 74.15Improved
set total diffusion timesteps T=50 instead of T=1 (ablation)TDSM · NTU-60 48/12 splittop-1 accuracy (%): 44.10 -> 56.03Improved
increase total timesteps from 50 to 500 (ablation)TDSM · NTU-60 55/5 splittop-1 accuracy (%): 86.49 -> 81.34Hurt
increase total timesteps from 50 to 100 (ablation)TDSM · NTU-60 48/12 splittop-1 accuracy (%): 56.03 -> 56.27No clear effect
use random Gaussian noise each training step instead of fixed noise (ablation)TDSM · NTU-60 55/5 splittop-1 accuracy (%): 76.40 -> 86.49Improved
use random Gaussian noise each training step instead of fixed noise (ablation)TDSM · NTU-120 96/24 splittop-1 accuracy (%): 52.21 -> 65.06Improved
use DiT backbone instead of U-Net (ablation)TDSM · NTU-60 48/12 splittop-1 accuracy (%): 51.12 -> 56.03Improved
predict skeleton feature z_x instead of noise (ablation)TDSM · NTU skeleton ZSARnoise-induced fluctuation in accuracy: fluctuation reduced by 5xMixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy